runggir: Run GGIR on the NIH biowulf.

In episphere/biowR: Helper functions for Biowulf Cluster

Run GGIR on the NIH biowulf.

Description

On the NIH biowulf, reading files directly from the /data disk is EXTREMELY slow. So we stage the data, files [f0..f1) on local disk and run all the local files.

The goal is to make a pipeline that is swarmable. Nothing is returned by this function, however, files are written in the results root.

The output of run_stage_one should would with the input of run_stage_2_5.

Usage

run_stage_one(cwa_root, results_root, json_args, f0, f1)

run_stages_2_5(part1_output_dir, json_args = "", f0, f1)

Arguments

cwa_root	where the CWA files are stored.
results_root	where to write the results
json_args	a json file containing a list of arguments passed into ggir
f0	the starting index of the files in cwa_root
f1	the ending index of the files in cwa_root
part1_output_dir	the output directory from running part 1 of the GGIR

Details

The output directory used in parts 2-5 is different from the results root from part 1. The output directory is inside the results root and has a directory path of RESULTS_ROOT/output_<studyname>

episphere/biowR documentation built on Dec. 6, 2023, 1:44 a.m.